Software Testing

Intermediate path · lesson 17 of 19

IntermediateNon-functional testing5 min readPublished November 12, 2026

Security testing for software testers

You don't need to be a hacker to find security bugs. Many serious flaws are found with a tester's mindset, two test accounts and the browser's developer tools. Here's what to check, the tools that help and the rules you must follow.

On this page

Key takeaways

  • Security testing checks that the system protects data and functions from people who shouldn't have them, including logged-in users who try to go beyond their rights.
  • Testers can find many serious flaws with everyday skills: access control checks, unexpected input, session handling and looking for leaked data.
  • Automated scanners, dependency checks and specialist penetration tests complement that. And always test only with permission.

Where this picks up

Security is a non-functional quality, and you've seen the most common API flaw in API testing. This article widens the view: what a tester, not a security specialist, can and should check.

Think like an attacker, test like a tester

Functional testing asks "does it work for the user?" Security testing adds: "what if the user is trying to misuse it?"

The OWASP Top 10, a widely used awareness list of the most critical web application risks, is a good map. In its 2021 edition, Broken Access Control is number one. Several other entries, like injection and vulnerable components, lead to the checks below.

What a tester can check

1. Access control

  • Log in as user B and open user A's URLs: /invoices/1001, /account/42/settings.
  • Call admin functions as a normal user, directly through the API if the button is hidden.
  • Change IDs, roles or prices in requests (developer tools, or an API tool).
  • Check that a hidden button isn't the only protection. The server must refuse too.

2. Input handling

Enter unexpected input in every field, URL parameter and header:

  • Very long strings, special characters, quotes (' " < >), HTML (<b>test</b>).
  • Text that looks like a database query or script.

Watch for errors that show internals (SQL messages, stack traces) and for input shown back on the page as HTML instead of plain text. Both are signs of injection or cross-site scripting risks worth reporting.

3. Authentication and sessions

  • Does logout really end the session? Press Back, or reuse the old session in another tab.
  • Do sessions expire after inactivity?
  • Is the account locked or slowed down after repeated failed logins?
  • Are password reset links single-use and time-limited? Can one user's reset link be used for another account?
  • Does changing the password end other active sessions?

4. Sensitive data

  • Does personal data appear in URLs, which end up in logs and browser history?
  • Do error messages reveal whether an email is registered ("no account with this email")?
  • Do API responses contain more data than the screen shows, like other users' details or internal fields?
  • Is everything served over HTTPS?

Tools and specialist testing

ApproachWhat it doesWho usually runs it
SAST (static analysis)Scans source code for insecure patternsPipeline, developers
SCA (dependency scanning)Finds libraries with known vulnerabilitiesPipeline
DAST (dynamic scanning)Attacks the running application automatically. Example: ZAP, an open-source scannerTesters, pipeline
Intercepting proxyLets you see and change requests between browser and server. Examples: ZAP, Burp SuiteTesters, security specialists
Penetration testingExpert, manual attack simulation within an agreed scopeSecurity specialists

Testers don't replace penetration testers. They catch common problems early and continuously, so specialists can focus on the hard ones.

Real-world example · Equifax, 2017A known flaw, a missing patch

In March 2017, a critical vulnerability in the Apache Struts web framework was disclosed, together with a fix. US authorities warned organisations about it the next day.

Equifax, one of the largest US credit reporting agencies, ran Struts in an online dispute portal. According to the US Government Accountability Office, the portal wasn't patched, and a scan meant to find vulnerable systems didn't identify it.

Attackers broke in from mid-May 2017 and stayed undetected for about 76 days. The personal data of around 147 million people was exposed.

Lesson: the vulnerability was public and the fix existed. What failed was knowing where the vulnerable component ran. Automated dependency scanning on every build, and an inventory of every system, are basic security testing.

Bringing it into everyday testing

  1. Keep two or more test users with different roles in every environment.
  2. Add access control checks to the test design for every feature that shows user data.
  3. Add a dependency scan to the pipeline and treat critical findings as release blockers.
  4. Run a DAST scan regularly against a test environment.
  5. Report security bugs privately and with high severity, following your organisation's process.

Do

Write security checks as normal test cases: "user B can't open user A's invoice through the URL or the API". They're part of the feature's correctness.

Don't

Assume the security team "does security". By the time a penetration test happens, the cheapest moment to find these bugs has passed.

Check yourself

After logging out, you press the browser's Back button and see your account page, and you can still change your address. What's wrong, and how serious is it?

Show answer

The session wasn't invalidated on logout: the server still accepts the old session. On a shared computer, the next person could use the account. That's a security bug, typically high severity. Report it with the steps, and check whether an old session token also still works through the API.